Measurement of WW Production with the ATLAS detector at s = 8 TeV
Kristin Lohwasser On behalf of the ATLAS collaboration
Theory
Event Selection Backgrounds Results
2
Measurements of WW Production
● Proves non-Abelian structure of electroweak interaction
→ ZWW vertex fixes unitarity problem
● Important test of the Standard Model
● Sensitive to new physics due to triple gauge couplings
● Previous measurement from LHC showed slight
enhancement of data over theoretical predictions
Theoretical background and motivation
1.4 1 1.7
3
Theoretical Predictions
MCFM NLO
contains triple gauge coupling
(enhanced cross section for high pT and new physics)
~91% ~2% ~7%
MCFM LO CT10 large k-factor?
NNLO MSTW2008 LHC Higgs XS WS
arXiv:1307.1347
Contributions neglected in this SM prediction
● qq → WW (NLO → NNLO+NNLL k-factor) + 1.6 pb
● gg → WW (LO → NNLO+NNLL k-factor) + 1.4 up to +2.8 pb
● Electroweak corrections – 0.5 pb
● γγ-induced WW + 0.5 pb
● Vector boson scattering + 0.5 pb
● Double parton interaction + 0.04 pb
Theoretical background and motivation
Total sum of:
+ 3.5 to 4.9 pb
Interference neglected
4
ATLAS Detector
Event Selection
muons and electrons
identified and well-isolated di- and single lepton trigger
pT>20 and 25 GeV
(subleading, leading)
Missing energy
vectorial sum of
calorimeter deposits
ETMiss,rel>45 GeV (ee,μμ) ETMiss,rel >15 GeV (eμ)
ETMiss,rel = ETMiss sin(ETMiss – jets,lep (if else ETMiss,rel = ETMiss)
Data taken at
s = 8 TeV
integrated luminosity
∫ L dt = 20.3 fb-1 4π-general purpose
detector
excellent particle ID and missing ET
performance
Missing momentum
vectorial sum of tracks pileup-robust
pTMiss> 45 and 20 GeV Δφ(pTMiss,ETMiss) <0.3/0.6
5
Event Selection
● Opposite sign requirement for leptons, outside Z mass window
● Top background dominates, rejected by Jet Veto requirement
● No jets with pT>25 GeV within ||<4.5
● Data driven determination of Z, top and W+jets events
eμ ee+μμ
Event Selection Event Selection
6
Data-driven Backgrounds
● Top background (12 – 14 %):
Jet veto efficiency applied to MC events to obtain normalization determined from a data control sample of top events with large scalar sum of lepton and jet pT
● W+jets background (eμ 4% / ee+μμ 1%):
extract normalization and shape
efficiencies for fake and real leptons for loose and tight cuts
determine signal and background components with matrix method
● Z+jets background (eμ 0.3% / ee+μμ 1%):
fit to data control region to extract normalization
Backgrounds
7
Backgrounds from Dibosons
● Diboson background determined using MC simulation WZ, ZZ, W and W*
● Validation region of same sign events with large W+jets contribution
eμ eμ
Missing ET Invariant dilepton mass m(ll)
Backgrounds
8
Distributions
● 6636 WW candidate events observed after full selection (5067 eμ, 594 ee, 975 μμ)
● WW MC signal scaled to measured cross sections
● Shapes well described
eμ eμ
eμ
scaled
scaled scaled
Results
9
Extraction of the cross section
Acceptance ee: 0.03 correction μμ: 0.04 eμ: 0.11
Branching ratio (0.108) (0.108) Total
Cross section
ETMiss reconstruction (2-4%)
Background (3-6 %)
Jet energy (2%)
Jet-veto requirement (4-5%)
Luminosity (2.5%)
Total: ~7.5%
Combination of channels performed using likelihood fit
Results
10
Combination
● Good agreement between the channels
● eμ dominates due to smaller uncertainty and larger statistics
Results
11
Results
● Some variations observed for different PDF predictions
● Global data CT10 PDF chosen for its generally good agreement with other LHC data
● PDF using LHC + HERA data only gives better agreement
Results
12
Conclusion
● WW cross section measured at s = 8 TeV using 20.3 fb-1
● Robust measurement with data-driven background estimations
● Result somewhat higher than prediction
● Some contributions neglected: Missing higher order corrections Could amount to several pb
Results public under ATLAS-CONF-2014-033
http://cds.cern.ch/record/1728248
